[QUOTE=Zartox;n186840]
Damn, looks like the e acute while editing a message make the message stop that character and erase the rest!
Will rewrite the main points:
First, Impressive!
Second: You load Weapons directory, rather than weapons directory (W case difference).
This make -weapon not to load on Linux because filesystem is case-sensitive.
Third: small glitch: dead sound not loading:
Offline mode succesfully initiated.
com.badlogic.gdx.utils.GdxRuntimeException: Error creating music com.badlogic.gdx.backends.lwjgl.audio.Wav$Music for file: /home/paul/T(e acute)l(e acute)chargements/graal_reborn/sounds/dead.wav
at com.badlogic.gdx.backends.lwjgl.audio.OpenALAudio.newMusic(OpenALAudio.java:130)
at com.badlogic.gdx.backends.lwjgl.audio.OpenALAudio.newMusic(OpenALAudio.java:41)
Four: works fine on Ubuntu 17.04, either with OpenJRE or OracleJRE.
[/QUOTE]
Thanks for reporting all of this, I’m surprised as much of it worked as it did on a linux system. There’s actually a linux build of libgdx I should compile with for your case which might fix the problem. In future formal releases I’ll do this.
So after having let this thread rest for a while, I can assure you I have not stopped with this project, however progress became slow moving to the networking as I’m very new to the scene and haven’t exactly decided how things should best be done. However… I’m partially done with networking for now, and although it’s somewhat crappy (or could be improved performance wise) I’ve decided to move to clientside features.
I now have a seteffect type function scripted, along with mechanics to distinguish certain npcs as “lights” and render them using a shader that accounts for the overall brightness of each pixel to determine its transparency in efforts to mimic what graal does.
Additionally, both images and animations can be rotated, scaled, and recoloured.
As for handling the hierarchy of which lights should be drawn, they’re drawn between the last ‘level’(tile) layer and the first ‘gui’ layer.
I also hope to draw all lights to the one frame buffer object and thus holes could be cut out for things like lantern effects.
What do you guys think of all this and is there features you’d like that I haven’t already stated?
Just going to start posting smaller more incremental progress.
So I’ve sorted out my animation rendering and now can display them as I would a normal image. The image shown is a result of a quick script I put together with a rotated copy of the npcs animation to follow the mouse and mimic a shadow.
Also sorry about the playback speed, my gif recorder(that I made) needs a bit of tweaking as it seems to be making the frames too short. If any of you could recommend some freeware you might use to capture gifs, that’d be nice.
[QUOTE=Dylan;n188165]unixmad is gonna try to hire you soon tricxta[/QUOTE]
I know this was meant as a joke, but it’s more likely than you might think.
It’s impossible to overstate how desperate unixmad is at this point. Unixmad has had zero programmers and zero people with skills since he broke up with Stefan over 3 years ago. PC Graal has had no updates and is being let die. Mobile cashcow versions barely stay online and are on the verge of irreversibly crashing where unixmad cannot fix by simply turning the server off and on again.
He is desperate and will rip off or exploit anyone he can get his hands on at this point.
It’s a great opportunity if you want to be paid in baguette mold then immediately fired, threatened and attacked after you complete the thing he wants to steal.
I am pretty sure unixmad and his few untalented brown-nosing admin staff watch these forums for opportunities to steal things.
[QUOTE=GOATSE;n188166]I am pretty sure unixmad and his few untalented brown-nosing admin staff watch these forums for opportunities to steal things.[/QUOTE]
Haven’t really being doing all that much with this project as of late due to the necessity of watching all the seasons of vikings but did a little work on implementing some showpoly functionality with the added bonus of transparency, scaling, and rotation being able to be defined. Wow so much better than graal.
So guys, one thing I haven’t mentioned is that I don’t have any support for “bigmap” type functionality, and instead I allow developers to choose the size of their levels (completely unrestricted). The reason behind this is not due to lazyness (although it is easier…) but I don’t feel like memory is an issue nowadays with most machines sporting 4gb+ ram, but this means I don’t have to have any kind of elaborate systems in place for baddies chasing you across a map or even a npc guiding you around (that’s not to say I don’t have support for npc warping amongst levels), but it’s not like it’s impossible to “merge” the levels into what would be a “gmap” where the whole map seems to be treating as a single entity and there’s no restrictions regarding moving npcs or players around… I don’t know…
But I want to know how you guys feel about this, what’s most preferable? From a collaborative sense, multiple fixed sized levels does seem to be the better option. So!.. Thoughts? Perhaps I should support both?
Multiple fixed size levels is good. On FTO we had Regions which you could determine the map width and height and all levels created on that region were sized correctly for that specific region. So, you still have variable level size, but with some constraints.
It was similar to a bigmap/gmap you determine the level width, height in tiles and then it handles creation, deleting and placement via the level editors UI. Quite similar to one of the example screenshots in Codr’s client thread
Also, by just having one giant map instead of sections of maps. Wouldn’t you encounter some issues with code being executed unnecessarily, when players are nowhere near. Audio triggers when players enter the map, but are nowhere near the trigger itself, timers starting, shit like that.
[QUOTE=Chicken;n188856]
Multiple fixed size levels is good. On FTO we had Regions which you could determine the map width and height and all levels created on that region were sized correctly for that specific region. So, you still have variable level size, but with some constraints.
It was similar to a bigmap/gmap you determine the level width, height in tiles and then it handles creation, deleting and placement via the level editors UI. Quite similar to one of the example screenshots in Codr’s client thread
Also, by just having one giant map instead of sections of maps. Wouldn’t you encounter some issues with code being executed unnecessarily, when players are nowhere near. Audio triggers when players enter the map, but are nowhere near the trigger itself, timers starting, shit like that.
[/QUOTE]
When you say you had multiple regions, you mean you had variable sized levels, but you give the “region” a grid x and y value and that’s where it sits on the overall map?
And yes there would be the potential for unnecessary execution of code which is why you as the developer would have to have a logic check for if the player is close enough to do anything, but the key point here is there isn’t a restriction as to when npc logic executes whereas you do have that restriction in bigmaps and gmaps where a npc might be sitting on the border of a neighbouring level and doing nothing taking away from the overall immersion. There’s nothing stopping me from introducing a player enters view event (similar to player enters) where its triggered each time the npc enters the player’s view based on the camera x,y and zoom.
Sounds about right. And when you wanted a new level, you could visually see where it would be on the region grid, when you created it. The regions also were able to handle levels that go in the z axis. We had insides/buildings in the positive Z axis and caves in the negative. Ended up with files like “RegionName 1, 0, -1.ext”. Sounds pretty complicated trying to explain it. However, for the end user it was excellent for map creating.
[QUOTE=Codr;n188863]Classic’s map is 21x58 levels, which translates to 4988928 tiles. Just throwing that out there. :)[/QUOTE]
So if a tile is represented by just an integer that’s 4 bytes right? So 4988928 x 4 bytes = ~20mb per layer, doesn’t seem too bad to me? Of course, with the number of npcs, there’s probably well over 1000 I’m guessing, and spread over so many levels, that’s to be expected and supposing each npc had an image or animation, yeah… things will probably get out of hand pretty quick…
I want to say people wouldn’t feel tempted to create such a massive map and have that as an expectation but I’m probably wrong. I do like chicken’s region idea but I don’t know how I would tackle loading when regions are large. For example, is a region loaded when you’re some arbitrary number of screens away from the region boundary?
